874 The American Naturalist. [October, 
The authof shows that the cleavage of willemite is like that of troost- 
ite, indistinct cleavages parallel to both the base and prism being 
made out in willemite. 
Composition of Staurolite and Arrangement of its Inclu- 
sions.—Exceptionally pure material for analysis was obtained by 
Penfield and Pratt’ from St. Gothard, Switz., Windham, Me., Lisbon, 
N. H., and near Burnsville, N. C. A powder of uniform specific grav- 
ity was obtained in each case by the use of fused silver nitrate as a 
separating fluid in a specially constructed apparatus, the heavier and 
lighter portions of the powder being in this way removed. Reckoning 
MnO and MgO as FeO, and Fe,O, as Al,O,, the four specimens yield 
results that agree well and indicate clearly that staurolite has the em- 
pirical formula H Al, Fe Si,O,, as already suggested by Groth. The 
silica alone does not agree closely with this formula, being in every 
case about one per cent too high, and the authors think that this is due to 
the presence of inclusions of quartz too minute to be separated from 
the powder. Carbonaceous inclusions are in the staurolite from Lis- 
The explanation of the authors is that the crystals of staurolite in 
growing in a solid rock, find it difficult to exclude foreign substances, 
the tendency to include them being greatest at the crystal edge and 
greatest where the interfacial angle is largest. 
Determination of Quartz and the Feldspars in thin Sec- 
tion.—Sometime since Becke described a method of distinguishing 
quartz from feldspar by treatment with hydrochloric acid and subse- 
quently tinting. He now" applies the same method to distinguish 
orthoclase from plagioclase and to determine the particular plagioclase 
species. Orthoclase is less affected by acid than plagioclase, and the 
soda rich plagioclases are less affected than the lime rich species. In 
rocks containing quartz, orthoclase and plagioclase, the slide is etched 
until by tinting the plagioclase shows an intense color. The orthoclase 
will then be faintly tinted and the quartz entirely unaffected 
Continuing his study Becke'* has devised methods for the same deter- 
minations based on differences of refractive index. The first method 
consists in the examination of a perpendicular contact plane between 
P Am. ime Sci., z xlvii, pp. 81-89. 
Tscherm. min, u. petrog. Mitth. = Heft 3, p. 2 (Notizen). 
MSitzungsber. d. k. Akad. d. Wissensch. i. Wien, Math. Naturw. Classe, Bd. 1, 
Abtk. I, pp. 358-376, July, 1893. 
